An individual tunable semiconductor laser can be used for spectroscopic analysis of solids, gases and liquids. Sometimes, the laser is optically coupled to an absorption cell, which serves two roles in the spectrometer system: (1) the absorption cell provides a defined absorption path length so that quantitative analysis of the sample can be implemented according to, for example, the Beer-Lambert relation for molecular absorption; and (2) the absorption cell can be configured to allow for greater optical path length than any single cell dimension through the use of multipass geometries. Greater optical path length translates to greater absorption signal, which in turn translates into a higher Signal to Noise Ratio (SNR) and a higher sensitivity.
Conventionally, when a tunable laser is used with an optical absorption cell, light from the laser is substantially collimated into the cell so as to reduce divergence and allow greater transmission through the cell and higher detection efficiency. This is true whether the light makes one single pass through the cell before detection or whether it traverses more than one pass as in Herriott cell or White cell based spectroscopy, cavity ringdown spectroscopy, and integrated cavity output spectroscopy (collectively: “multipass spectroscopy”). For more on integrated cavity output spectroscopy, see U.S. Pat. No. 6,795,190, entitled “Absorption Spectroscopy Instrument with Off-Axis Light Insertion into Cavity,” which is incorporated herein by reference in its entirety.
Not all spectroscopy systems use absorption cells. For example, the laser beam may be directed toward a target and its wavelength tuned while the backscattered photons are detected and analyzed. A common instance of this embodiment is the standoff detection of surface adsorbed condensed phase material using, for example, infrared absorption or Raman spectroscopies.
For broader wavelength coverage tunable laser arrays can also be used for spectroscopy, both with and without absorption cells. For more on continuously or broadly tunable laser arrays, see U.S. Pat. No. 7,826,509 “Broadly Tunable Single-Mode Quantum Cascade Laser Sources and Sensors,” which is incorporated herein by reference in its entirety. In this work, an array of single frequency Quantum Cascade Lasers, each member with its own wavelength, is used to obtain broader absorption spectra than would be obtainable using just one single frequency laser. This is especially useful for the spectroscopic analysis of condensed phased materials and for analysis of multiple gases, where broader spectral coverage is highly beneficial.
In either case, using a single emitter or an array of emitters, the laser beams' divergence often limits the obtainable signal to noise ratio as high divergence reduces the available optical transmitted or backscattered optical power per unit area for detection when interrogating samples over long path lengths, including those in multipass sample cells or standoff spectroscopy. In addition, portions of the diverging beams may interrogate different portions of the sample volume. As a result, different portions of the detected absorption spectrum may correspond to different parts of the sample. This can be especially troublesome when interrogating condensed phase samples that are not perfectly homogenous.
In the case of single emitters, the common solution to reduce divergence is to place an optical element with focusing power, such as a lens, in front of the emitter. However, for emitter arrays, a single lens is not always appropriate. This is illustrated in FIG. 1, which illustrates how diverging beams 111 emitted by lasers 112 in a laser array 110 are coupled to a single lens 120. The lens 120 is positioned with its optical axis perpendicular to the face of the laser array 120 at a distance roughly equal to the lens's focal length and produces a set of collimated beams 121. Although the lens 120 collimates the diverging beams 111 emitted by the laser array, the resulting collimated beams 121 are tilted with respect to each other by an amount that depends on both the beam's position. If the lens 120 has a focal length f, the collimated beam 121 from the nth laser in the array 110 points at an angle given by:
                              Δ          ⁢                                          ⁢                      θ            n                          =                              tan                          -              1                                ⁡                      (                                          Δ                ⁢                                                                  ⁢                                  x                  n                                            f                        )                                              (        1        )            
In equation (1), Δθn is measured with respect to the axis of the lens 120 and Δxn is the transverse position of the nth laser in the array 110 relative to the focal point of the lens 120. In this case, the laser beams are spatially separated in the far-field 130, as shown. Thus, collimation of diverging beams 111 from an entire array 110 using a single lens 120 exacerbates, rather than corrects, many of the problems associated with using a tunable laser array instead of a single tunable laser for spectroscopy by introducing an angular difference between output beams where previously there was only a lateral offset.
Wavelength beam-combining (WBC) (also known as spectral beam combining) is a technique used in telecommunications and spectroscopy to address the problem of position- and wavelength-dependent steering caused by collimating beams from an array. In WBC, the laser beams from an array are merged into a single, co-linear beam of spatially overlapping outputs that can then be propagated into the far field or perhaps through an absorption cell. This method takes on many physical embodiments, depending on the application.
For example, in Wavelength Division Multiplexing, a central technique in telecommunications, the outputs of multiple fiber-coupled single frequency lasers are merged into one single fiber for long distance transmission.
Yet another example is the WBC of free space infrared lasers using either “open loop” or “closed loop” approaches. One example of open loop beam combining involves an array of single frequency QCLs that are coupled to a non-actively aligned dispersion element using either an array of matched lenses or a single collection lens. The blaze spacing and angle of the grating are chosen to correct for the physical and wavelength separation of the array members such that the grating disperses each laser wavelength at a slightly different angle. The result is that all wavelengths are overlapped in space and propagated together (co-linearly) in the far field.
In the closed loop approach to WBC, the array instead comprises more-or-less identical emitters, where the emission wavelength is not differentiated in the laser itself. Rather, the dispersion element, in combination with an output coupler, acts to form a laser cavity such that each laser's frequency is determined by feedback from the dispersion element. As with the “open loop” approach, this technique results in a beam of spatially overlapping contributions of different frequencies.
In both open-loop and closed-loop WBC, a beam-combined laser array system has some advantages over non-beam combined array outputs, including higher degree of spatial symmetry and an improved beam quality parameter, M2. Indeed, beam combining of QCL arrays has be shown to produce near diffraction limited (where M2=1) performance in the far-field. Improved beam symmetry, spatial overlap, and M2 are all useful in many applications in molecular spectroscopy, infrared countermeasures (IRCM), and fiber coupling of laser arrays.